August 4, 2026

expert reaction to comparison of NICE criteria with the BOADICEA multifactorial risk model to guide breast cancer risk assessment and referral amongst women under age 50

A study published in the British Journal of Cancer compares the NICE criteria with the BOADICEA risk model for breast cancer risk assessment and referral.

Prof Paul Pharoah, Professor of Cancer Epidemiology, Cedars-Sinai Health Sciences University, said:

“The science is this study is of high quality.  The researchers have used a large cohort of women under the age of 50 and calculated their future risk of developing cancer at the time of recruitment into the study using a well-established breast cancer risk prediction model called BOADICEA.The women were then followed for 10 years to find out who did and did not develop cancer.  The BOADICEA risk model performed well based on standard metrics for assessing the performance of a risk model.

“The researchers then used the referral criteria from guidelines developed by NICE to evaluate what proportion of women that were identified as being at high risk by BOADICEA would be eligible for referral for breast cancer risk management.  The NICE referral criteria are based on the number of relatives in a woman’s family who have had breast cancer (this is known as a family history of breast cancer).  The researchers also estimated the proportion of the women who went on to get breast cancer in the next ten years that were eligible for referral.  Less than 5% of high-risk women and less than 5% of future cases were found to be eligible for referral.  However, the fundamental premise to justify these analyses is flawed and the interpretation of the results is misleading.

“The press release starts by stating “The National Institute for Health and Care Excellence (NICE) criteria used to decide who should be referred by GPs for further breast cancer risk assessment and specialist care misses up to 95% of women under 50 who will go on to develop the disease in the next 10 years.”  This implies that the purpose of the NICE guidelines is to identify women at high risk.It is not.  The NICE guidelines are intended to guide the management of women who present to their GP because they are worried about their family history of breast cancer.Most women do not have a family history of breast cancer and so the guidelines are not relevant for those women.  The fact that most women aged 20 to 50 who go on to develop breast cancer in the next ten years are not identified by the NICE guidelines criteria is an expected finding – 80% of breast cancers in this group of women occur in women without any family history of breast cancer.  Similarly, most women at high-risk do not have a family history and criteria intended to help manage women who are worried about their family history will inevitably perform poorly at a different task.

“Finally, Dr Usher-Smith states “We need to get better at identifying women at highest risk of breast cancer so that we can intervene early, when there are more options for treating, or even preventing, their disease.”  It is unclear how identifying women under the age of 50 who are high risk enable early intervention to prevent or treat disease.  However, there are no good data to show that any intervention, such as mammographic screening, in these women improves outcomes.  There is no good rationale for widespread risk prediction in these women.

Dr Adam Brentnall, Reader in Biostatistics, Queen Mary University of London (QMUL), said:

“This is good quality research, using data from the Breast Cancer Now Generations Study. This study recruited more than 100 thousand volunteers in the UK without breast cancer. When they joined participants filled in questionnaires and gave samples that could be used for genetic testing, which was used for risk assessment in this study. A subset of the participants was used for analysis, including those who were later diagnosed with breast cancer.

“The study demonstrates the substantial value of comprehensive breast cancer risk assessment over current criteria recommended to identify women at increased risk. However, it is worth noting some uncertainty in the statistics used in the press release. For instance, based on the statistical analysis reported, the percentage of women who would be missed with the NICE referral criteria would technically be up to ~92%, to account for the confidence intervals. 

“Further, as with all studies with volunteers, it is important that everyone is represented in adequate numbers to enable conclusions to be drawn about the wider population. While the authors try to address potential selection effects, there remains a risk that this was imperfectly done. 

“A more fundamental limitation is that everyone included in the analysis is white. Therefore, clearly more work is needed to understand the implications of comprehensive risk assessment, including use of genetic testing, across all the full ethnic diversity of the UK population.

“Another limitation of the study is that the model examined does not include all known risk factors for breast cancer. This includes breast density, which is of similar predictive value to polygenic risk scores [1]. There is also increasing evidence that artificial intelligence (AI) algorithms using imaging can evaluate breast cancer risk much better than traditional risk models like BOADICEA or Tyrer-Cuzick [2,3], including if polygenic risk scores are used [4].

“Therefore, the paper demonstrates proof of concept that risk assessment including a questionnaire and polygenic risk scores could be a very useful approach, at least statistically. But the risk assessment could be made even better.

“The findings fit with existing evidence, including some earlier work reported more than a decade ago using the IBIS (Tyrer-Cuzick) model in a Manchester cohort [5] – which also showed the benefit of risk assessment vs NICE criteria. One way the results extended the previous work on NICE criteria is by considering the implication of a certain type of genetic test (polygenic risk score). This substantially increased the number of (white) women identified at moderate risk, vs a questionnaire alone.

“NICE guideline CG164 recommends offering annual mammographic screening to all females aged 40 to 49 years at moderate risk of breast cancer. GPs are instructed to refer to a specialist clinic for risk assessment based on certain criteria. However, these criteria miss most women who are actually at moderate risk when you use a model like BOADICEA, which often gets used after a woman has been referred by her GP.

“There are NICE guidelines for screening women in their 40s based on risk. However, most women in their 40s who would qualify for screening do not get screened because they don’t know their risk. If routine risk assessment was offered proactively, then many more women at increased risk could be screened in their 40s.

“There is direct evidence this would improve outcomes. This includes results from a UK randomised trial that showed annual mammography screening women in their 40s reduces breast cancer deaths [6].”

[1] van Veen EM, Brentnall AR, Byers H, Harkness EF, Astley SM, Sampson S, Howell A, Newman WG, Cuzick J, Evans DGR. Use of Single-Nucleotide Polymorphisms and Mammographic Density Plus Classic Risk Factors for Breast Cancer Risk Prediction. JAMA Oncol. 2018 Apr 1;4(4):476-482. doi: 10.1001/jamaoncol.2017.4881. PMID: 29346471; PMCID: PMC5885189.
[2] Damiani C, Kalliatakis G, Sreenivas M, Al-Attar M, Rose J, Pudney C, Lane EF, Cuzick J, Montana G, Brentnall AR. Evaluation of an AI Model to Assess Future Breast Cancer Risk. Radiology. 2023 Jun;307(5):e222679. doi: 10.1148/radiol.222679. PMID: 37310244.
[3] Eriksson M, Czene K, Scott C, Stoddard S, Smith H, Hall P, Vachon C. A long-term image-derived AI-based risk model for primary prevention of breast cancer in individuals at high risk. Sci Transl Med. 2026 May 20;18(850):eady7414. doi: 10.1126/scitranslmed.ady7414. Epub 2026 May 20. PMID: 42160452.
[4] Medha Kaul, Christopher G Scott, Alena Wadzinske, Imon Banerjee, Dan Hursh, Aaron Norman, Fang-Fang Wu, Ramon Correa-Medero, Fergus Couch, Sandhya Pruthi, Despina Kontos, Anne Marie McCarthy, Janet Olson, Stacey Winham, Karla Kerlikowske, Celine M Vachon, Performance of clinical breast cancer risk prediction models vs a mammography-based artificial intelligence risk model, JNCI: Journal of the National Cancer Institute, 2026;, djag083, https://doi.org/10.1093/jnci/djag083<https://doi.org/10.1093/jnci/djag083>
[5] Evans DG, Brentnall AR, Harvie M, Dawe S, Sergeant JC, Stavrinos P, Astley S, Wilson M, Ainsworth J, Cuzick J, Buchan I, Donnelly LS, Howell A. Breast cancer risk in young women in the national breast screening programme: implications for applying NICE guidelines for additional screening and chemoprevention. Cancer Prev Res (Phila). 2014 Oct;7(10):993-1001. doi: 10.1158/1940-6207.CAPR-14-0037. Epub 2014 Jul 21. PMID: 25047362.
[6] Duffy SW, Vulkan D, Cuckle H, Parmar D, Sheikh S, Smith RA, Evans A, Blyuss O, Johns L, Ellis IO, Myles J, Sasieni PD, Moss SM. Effect of mammographic screening from age 40 years on breast cancer mortality (UK Age trial): final results of a randomised, controlled trial. Lancet Oncol. 2020 Sep;21(9):1165-1172. doi: 10.1016/S1470-2045(20)30398-3. Epub 2020 Aug 12. PMID: 32800099; PMCID: PMC7491203.

‘Comparison of NICE criteria with the BOADICEA multifactorial risk model to guide breast cancer risk assessment and referral amongst women under age 50 within primary care by Reuben Frost et al. was published in British Journal of Cancer at 02:00 UK Time on Tuesday 4 August 2026. 

DOI: 10.1038/s41416-026-03547-2

Declared interests

Prof Paul Pharoah: “I have no conflicts of interest to declare.”

Dr Adam Brentnall: “I receive royalty payments from Cancer Research UK arising from commercial use of the IBIS (Tyrer-Cuzick) breast cancer risk evaluation tool. I have received research funding from Cancer Research UK, Breast Cancer Now, MRC for research on breast cancer risk. I am a member of the UK National Screening Committee research and methodology subgroup, and other potential conflicts are listed here https://www.gov.uk/government/publications/uk-national-screening-committee-register-of-interests/research-and-methodology-group”